Which metric measures the model's ability to detect positive instances among all actual positives?

• A. Recall
• B. Accuracy
• C. Precision
• D. F-score

Answer: (A)

Detecting all actual positives is about sensitivity. Recall measures the model's ability to identify positive instances when they are truly positive. It is calculated as the number of true positives divided by the sum of true positives and false negatives. This focus on actual positives makes recall the right choice when the goal is to capture as many positives as possible, such as screening tests, where missing a positive can have serious consequences.

Other metrics describe different aspects: accuracy looks at overall correctness across both classes and can be misleading if the classes are imbalanced; precision measures how many of the predicted positives are actually positive, focusing on the quality of positive predictions rather than coverage; the F-score combines precision and recall, but the question targets the ability to detect positives among all actual positives, which is precisely what recall evaluates.